Reproduce “Numerical Embeddings for CTR Prediction: A Taxonomy, Unified Benchmark, and Open-Source FuxiCTR Extension” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Heng Yao, Hanwei Zhu, Tianying Liu, Yulou Shu, Siyun Hou
The authors introduce NumEmbBench, a benchmark comparing 18 numerical-embedding configurations through a shared interface across three datasets and nine backbones for click-through-rate prediction. Under train-only z-score preprocessing, DEER has the best mean AUC rank, 3.41, with 14 wins in 27 cells. It is not a universal winner: frozen raw-scale evaluation changes the winner in 11 cells.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.